Abstract
Background
Emactuzumab is a monoclonal antibody against the colony-stimulating factor-1 receptor and targets tumor-associated macrophages (TAMs). This study assessed the safety, clinical activity, pharmacokinetics (PK) and pharmacodynamics (PD) of emactuzumab, as monotherapy and in combination with paclitaxel, in patients with advanced solid tumors.
Patients and methods
This open-label, phase Ia/b study comprised two parts (dose escalation and dose expansion), each containing two arms (emactuzumab, every 2 or 3 weeks, as monotherapy or in combination with paclitaxel 80 mg/m2 weekly). The dose-escalation part explored the maximum tolerated dose and optimal biological dose (OBD). The dose-expansion part extended the safety assessment and investigated the objective response rate. A PK/PD analysis of serial blood, skin and tumor biopsies was used to explore proof of mechanism and confirm the OBD.
Results
No maximum tolerated dose was reached in either study arm, and the safety profile of emactuzumab alone and in combination does not appear to preclude its use. No patients receiving emactuzumab monotherapy showed an objective response; the objective response rate for emactuzumab in combination with paclitaxel was 7% across all doses. Skin macrophages rather than peripheral blood monocytes or circulating colony-stimulating factor-1 were identified as an optimal surrogate PD marker to select the OBD. Emactuzumab treatment alone and in combination with paclitaxel resulted in a plateau of immunosuppressive TAM reduction at the OBD of 1000 mg administered every 2 weeks.
Conclusions
Emactuzumab showed specific reduction of immunosuppressive TAMs at the OBD in both treatment arms but did not result in clinically relevant antitumor activity alone or in combination with paclitaxel. (ClinicalTrials.gov Identifier: NCT01494688)
Keywords: CSF-1R, tumor-associated macrophages, emactuzumab, paclitaxel, tumor microenvironment
Key Message
Anti-CSF-1R emactuzumab targeted and depleted M2-like immunosuppressive tumor-associated macrophages in patients with solid malignancies. Treatment provided evidence of the tolerability of emactuzumab as monotherapy and in combination with paclitaxel. Neither emactuzumab monotherapy nor combination with chemotherapy resulted in clinically relevant antitumor activity.
Introduction
Macrophages are part of the mononuclear phagocyte system and demonstrate high phenotypic and functional plasticity in vivo. The phenotypic extremes can be described as macrophages that are involved in the inflammatory response, pathogen clearance, as well as antitumor immunity (M1-like subtype) or that inhibit the anti-inflammatory response, wound healing and have pro-tumorigenic properties (M2-like subtype) [1, 2].
Colony-stimulating factor (CSF-1)-mediated signaling via its receptor (CSF-1R) is a key regulator of the proliferation, migration, survival and differentiation of macrophages and their precursors [3]. CSF-1 is highly expressed by several tumors and associated with the presence of tumor-associated macrophages (TAMs) that closely resemble the M2-polarized functional phenotype. The presence of TAM appears to be an adverse prognostic factor associated with poor survival in patients with solid tumors and hematological malignancies [4]. Targeting CSF-1R signaling in TAMs may therefore be a promising therapeutic approach by inhibiting tumor-promoting macrophages in the tumor microenvironment (TME).
Emactuzumab is a humanized mAb directed against CSF-1R expressed on macrophages [5, 6] and has demonstrated a profound antitumor effect through interference with the CSF-1/CSF-1R axis, along with a manageable safety profile in patients with diffuse-type giant cell tumor (dt-GCT) [5, 7]. Here, we report on the safety, clinical activity, pharmacokinetics (PK) and pharmacodynamics (PD) of emactuzumab, both as monotherapy and in combination with paclitaxel, in patients with advanced/metastatic solid tumors.
Patients and methods
Study design
This open-label, multicenter dose escalation and expansion phase Ia/b study (NCT01494688) contained two arms (emactuzumab alone or in combination with paclitaxel). Primary objectives were to evaluate the safety, tolerability and PK and to determine the maximum tolerated dose (MTD) and/or optimal biological dose (OBD). Secondary objectives were to investigate PD effects in tumor and surrogate tissues, PD effects based on [18F]2-fluoro-2-deoxy-D-glucose (FDG–PET) and any preliminary clinical activity.
Study treatments
In the dose-escalation part with emactuzumab alone, patients received doses ranging from 100 to 3000 mg i.v. administered over 1.5 h q2w in a 3 + 3 design. For patients receiving emactuzumab in combination with paclitaxel, the starting dose was 200 mg (based on safety as monotherapy) and increased to 2000 mg i.v. administered over 1.5 h. Paclitaxel was given as a fixed weekly dose of 80 mg/m2 i.v. The study design included a 3-week screening phase, a treatment phase and a 4-week follow-up phase per patient.
In the dose-expansion part of the study, patients were treated with emactuzumab at the established OBD of 1000 mg q2w alone or in combination with paclitaxel. A separate small cohort of five patients was treated with an exploratory dose schedule of emactuzumab 1350 mg q3w alone. All PK/PD data were derived from patients treated with the q2w dose schedule only. Treatment was continued until disease progression or dose-limiting toxicity (DLT), any unacceptable toxicity or until patient consent withdrawal.
Patients
Eligible patients had histologically confirmed advanced and/or metastatic solid tumors (for dose escalation with paclitaxel combination: only ovarian and breast cancer patient; for the expansions cohorts: only patients with dt-GCT, soft tissue sarcoma or malignant mesothelioma, ovarian, endometrial and breast cancer, and pancreatic cancer), were ≥18 years of age, had an Eastern Cooperative Oncology Group performance status of ≤1, had measurable disease according to Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1, and had adequate bone marrow, cardiac, liver and renal function. The study was conducted in accordance with the Declaration of Helsinki and the International Conference on Harmonisation guidelines for good clinical practice. The protocol was approved by the ethics committees of participating centers.
Assessments
Adverse events (AEs) were monitored and graded according to National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.03. [8, 9]. A DLT was any related AE grade ≥3 in severity with exceptions as defined in the supplementary Material, available at Annals of Oncology online. The MTD was the highest dose cohort evaluated within which < 33% of patients experienced a DLT.
Centrally reviewed radiologic response was assessed with computed tomography scan and/or MRI at baseline and after every 6 weeks. Disease response was defined using RECIST 1.1 [10]. Metabolic response, based on FDG–PET imaging, was assessed at baseline and after 4 weeks of treatment. FDG–PET imaging assessment was based on European Organization for Research and Treatment of Cancer criteria [11].
Pre-treatment and on-treatment blood samples for PK, safety and PD assessments were taken in both parts of the study. Noncompartmental analysis was conducted on emactuzumab serum concentration data.
Biopsies at baseline and after 2 weeks (skin) and 4 weeks (tumor) of treatment were taken, fixed in 10% buffered formalin and embedded in paraffin. Immunohistochemical (IHC) analysis was carried out using commercial antibodies with automated Benchmark XT system (all from Ventana Medical Systems, Tucson, AZ). Monocyte subsets in the peripheral blood were monitored using flow cytometry (Becton Dickinson, Frnklin Lakes, USA). CSF-1 protein isoforms were quantified using the Elecsys® (Roche Diagnostics, Mannheim, Germany) CSF-1 assay operated on the e601 module of the cobas® 6000 system (Roche Diagnostics, Mannheim, Germany). RNA sequencing was carried out from bulk paired tumor biopsy tissue using Illumina TruSeq RNA Access Library Preparation and instrumentation. Further details on carried out assays are provided in the supplementary Material, available at Annals of Oncology online!.
Results
Patients
Altogether, 99 patients with solid tumors were enrolled to treatment with emactuzumab alone (29 patients enrolled into eight dose cohorts; 70 patients enrolled into expansion) (Table 1 and supplementary Figure S1, available at Annals of Oncology online). In total, 88 patients discontinued the study because of progressive disease, one patient discontinued because of death from progressive disease, eight patients were withdrawn because of an AE (see Table 2 footnotes for details), one patient was withdrawn because of noncompliance and one patient withdrew consent.
Table 1.
Patient demographics and baseline characteristics
| Characteristics | Monotherapy | Combination with paclitaxel |
|---|---|---|
| N = 99a | N = 54a | |
| Age, years | ||
| Median | 63 | 53 |
| Range | 18–80 | 30–77 |
| Sex | ||
| Female | 49 (49) | 53 (98) |
| Male | 50 (51) | 1 (2) |
| ECOG performance status | ||
| 0 | 38 (38) | 25 (46) |
| 1 | 61 (62) | 29 (54) |
| No. of prior treatment regimens (including adjuvant) | ||
| <3 | 37 (37) | 12 (22) |
| ≥3 | 62 (63) | 42 (78) |
| Site of primary cancer | ||
| Mesothelium (pleura + peritoneum) | 15 (15) | – |
| Ovary | 15 (15) | 13 (24) |
| Pancreas | 15 (15) | – |
| Colorectal | 11 (11) | – |
| Soft tissue | 10 (10) | – |
| Uterus | 7 (7) | – |
| Breastb | 5 (5) | 40 (74) |
| Other | 21 (21) | 1 (2)c |
| Dose levels: dose escalation | ||
| 100 mg q2w | 1 | NA |
| 200 mg q2w | 1 | 4 |
| 400 mg q2w | 7 | 4 |
| 600 mg q2w | 5 | 5 |
| 900 mg q2w | 3 | 6 |
| 1350 mg q2w | 2 | 7 |
| 2000 mg q2w | 4 | 8 |
| 3000 mg q2w | 6 | – |
| Dose levels: cohort expansion | ||
| 1000 mg q2w | 65 | 20 |
| 1350 mg q3w | 5 | – |
Data are presented as N (%).
Included in the safety and efficacy analysis are all patients who received at least one dose of emactuzumab and had measurable disease at baseline.
The 20 patients with breast cancer from the dose-expansion part were all HER2 negative.
Fallopian tube.
ECOG, Eastern Cooperative Oncology Group; NA, not applicable; q2w, every 2 weeks; q3w, every 3 weeks.
Table 2.
Summary of adverse events (AEs) occurring in ≥10% of patients across all dose levels (i.e. 100–3000 mg for monotherapy and 200–1350 mg for combination with paclitaxel)
| Monotherapya |
Combination with paclitaxela |
|||||||
|---|---|---|---|---|---|---|---|---|
|
N = 99 |
N = 54 |
|||||||
| Irrespective of relationship |
Related |
Irrespective of relationship |
Related |
|||||
| AEs | Any grade | Grade ≥3 | Any grade | Grade ≥3 | Any grade | Grade ≥3 | Any grade | Grade ≥3 |
| Any AE | 96 (97) | 51 (52) | 76 (77) | 14 (14) | 54 (100) | 37 (69) | 52 (96) | 26 (48) |
| AEs leading to treatment discontinuationb | 8 (8) | 6 (6) | 4 (4) | 2 (2) | 8 (15) | 6 (11) | 5 (9) | 3 (6) |
| Asthenia | 49 (49) | 7 (7) | 27 (27) | 4 (4) | 42 (78) | 5 (9) | 35 (65) | 4 (7) |
| Anemia | 31 (31) | 9 (9) | 9 (9) | 3 (3) | 28 (52) | 10 (19) | 11 (20) | 4 (7) |
| Decreased appetite | 31 (31) | 1 (1) | 11 (11) | – | 18 (33) | 1 (2) | 10 (19) | – |
| Periorbital edema | 29 (29) | – | 29 (29) | – | 29 (54) | – | 28 (52) | – |
| Peripheral edema | 25 (25) | – | 18 (18) | – | 18 (33) | 1 (2) | 15 (28) | 1 (2) |
| Vomiting | 23 (23) | 1 (1) | 3 (3) | – | 19 (35) | – | 10 (19) | – |
| Pyrexia | 22 (22) | – | 7 (7) | – | 22 (41) | – | 6 (11) | – |
| Pruritus | 21 (21) | – | 18 (18) | – | 10 (19) | 1 (2) | 8 (15) | 1 (2) |
| Nausea | 21 (21) | 1 (1) | 7 (7) | 1 (1) | 24 (44) | – | 15 (28) | – |
| Abdominal pain | 20 (20) | 3 (3) | 4 (4) | – | 9 (17) | 1 (2) | 4 (7) | – |
| Diarrhea | 19 (19) | – | 8 (8) | – | 20 (37) | – | 11 (20) | – |
| Constipation | 18 (18) | – | 1 (1) | – | 15 (28) | 1 (2) | 4 (7) | 1 (2) |
| Dyspnea | 17 (17) | 3 (3) | – | – | 23 (43) | 2 (4) | 6 (11) | – |
| Eyelid edema | 16 (16) | – | 14 (14) | – | 6 (11) | – | 6 (11) | – |
| Fatigue | 16 (16) | 3 (3) | 11 (11) | 2 (2) | 6 (11) | 4 (7) | 5 (9) | 4 (7) |
| Face edema | 15 (15) | – | 14 (14) | – | 14 (26) | – | 14 (26) | – |
| Hypertension | 11 (11) | 2 (2) | 1 (1) | 1 (1) | 6 (11) | 1 (2) | 3 (6) | 1 (2) |
| Headache | 10 (10) | – | 5 (5) | – | 13 (24) | – | 2 (4) | – |
| Upper abdominal pain | 9 (9) | – | 3 (3) | – | 9 (17) | 1 (2) | 2 (4) | – |
| Cough | 9 (9) | – | – | – | 7 (13) | – | 1 (2) | – |
| Anxiety | 9 (9) | – | 1 (1) | – | 7 (13) | – | – | – |
| Rash | 9 (9) | – | 9 (9) | – | 6 (11) | 1 (2) | 4 (7) | 1 (2) |
| Insomnia | 7 (7) | – | – | – | 6 (11) | – | – | – |
| Epistaxis | 5 (5) | – | 3 (3) | – | 13 (24) | – | 10 (19) | – |
| Myalgia | 5 (5) | – | 3 (3) | – | 9 (17) | – | 7 (13) | – |
| Infection, device related | 4 (4) | – | – | – | 8 (15) | 1 (2) | – | – |
| Infection, oral fungal | 4 (4) | – | 1 (1) | – | 7 (13) | – | 2 (4) | – |
| Dysgeusia | 4 (4) | – | 3 (3) | – | 7 (13) | – | 5 (9) | – |
| Mucosal inflammation | 4 (4) | – | 4 (4) | – | 6 (11) | – | 5 (9) | – |
| Hypophosphatemia | 3 (3) | – | 3 (3) | 2 (2) | 8 (15) | 8 (15) | 5 (9) | 5 (9) |
| Peripheral neuropathy | 2 (2) | – | 2 (2) | – | 12 (22) | 1 (2) | 7 (13) | 1 (2) |
| Hypocalcemia | 2 (2) | – | 1 (1) | – | 6 (11) | 3 (6) | 3 (6) | 1 (2) |
| Paraesthesia | – | – | – | – | 11 (20) | – | 4 (7) | – |
| Alopecia | – | – | – | – | 7 (13) | – | 6 (11) | – |
Data are presented as N (%).
Percentages were calculated using the number of patients in the safety population as denominator. Multiple occurrences of the same AE in one individual counted only once.
AEs leading to treatment discontinuation in the emactuzumab monotherapy arm were single patients with asthenia (grade 2, related), cerebral ischemia (grade 5, unrelated), increased blood creatine phosphokinase (grade 4, related) and myalgia (grade 1, unrelated) occurring in one patient, depression (grade 3, unrelated), hematoma (grade 4, related), increased intraocular pressure (grade 1, unrelated), laryngeal edema (grade 2, related), and lung infection (grade 4, unrelated).
SAE, serious adverse event.
In total, 54 patients were enrolled to treatment with emactuzumab in combination with paclitaxel (34 patients enrolled into six dose cohorts; 20 patients enrolled into expansion). A total of 42 patients discontinued the study because of progressive disease, two patients discontinued because of death (one caused by progressive disease and one caused by gastrointestinal perforation, considered a DLT), eight patients were withdrawn because of AEs (see Table 2 footnotes for details) and two patients withdrew for other reasons.
Safety
Dose levels were escalated up to 3000 mg for monotherapy and 2000 mg in combination with paclitaxel, and no MTD was reached.
No DLTs were recorded up to the highest dose tested for monotherapy. Asthenia, anemia and decreased appetite occurred most often (>30% of patients) (Table 2). Overall, 52% of patients receiving emactuzumab alone experienced grade 3/4 AEs, with only asthenia and anemia occurring in >5% of patients.
Two patients treated with emactuzumab and paclitaxel experienced DLTs (one patient at 1350 mg (grade 4 hypokalemia and grade 3 gastrointestinal inflammation and hemorrhagic enterocolitis) and one patient at 2000 mg (grade 5 gastrointestinal perforation), the latter being the only patient who died because of a treatment-related AE). Asthenia, periorbital edema and anemia occurred most often (>50% of patients) (Table 2). Overall, 69% of patients receiving emactuzumab and paclitaxel experienced grade 3/4 AEs, with only anemia and hypophosphatemia occurring in >10% of patients.
Shifts in serum enzyme levels were seen in patients and are consistent with delayed clearance of serum enzymes secondary to emactuzumab-induced Kupffer cell depletion [12]. The vast majority of patients remained asymptomatic despite continued treatment at the recommended dose. None of the patients developed liver failure or rhabdomyolysis (supplementary Material, available at Annals of Oncology online).
Clinical activity
No patients receiving emactuzumab monotherapy showed an objective response, and 13 patients (13%) had stable disease. In patients receiving emactuzumab in combination with paclitaxel, four patients showed a confirmed partial response, two of whom had received prior taxane treatment of metastatic disease. Three of the responding patients had estrogen receptor-positive and HER2-negative breast cancer, and one patient had ovarian cancer. Altogether, 23 patients (43%) had stable disease.
For FDG–PET, in the monotherapy arm, partial metabolic response was observed in 11 patients (11%), and 40 patients (40%) had stable metabolic disease. In the paclitaxel combination arm, partial metabolic response was observed in 21 patients (39%), and 16 patients (30%) had stable metabolic disease.
Pharmacokinetics and pharmacodynamics
The PK of emactuzumab were nonlinear from 100 to 900 mg. For doses >900 mg, exposure increased approximately dose proportionally, indicative of >90% target saturation.
We could confirm that the main target cell population in the peripheral blood were the nonclassical CD14DIMCD16BRIGHT monocytes whereas other monocyte subsets showed less or no susceptibility to emactuzumab treatment according to their described CSF-1R expression (supplementary Figure S2A, available at Annals of Oncology online). The functional properties of monocytes are discussed controversially. Non-classical monocytes have been reported as the actual progenitor of M2-like macrophages as they seem to be involved in tissue repair [13] as well as displaying inflammatory characteristics such as interleukin 13 and tumor necrosis factor-α production upon activation [14]. When we assessed the change from baseline of the intermediate CD14BRIGHTCD16DIM monocytes in relation to the steady-state concentration of emactuzumab at different dose levels, we observed that a nadir for the monocyte reduction in the peripheral blood was reached at exposures correlating to dose levels of emactuzumab ∼400 mg from cycle 2 onward for both monotherapy and combination therapy with paclitaxel (Figure 1A).
Figure 1.
Pharmacodynamic markers to support optimal biological dose selection in patients treated with emactuzumab monotherapy. (A) Nonclassical monocyte levels (cells/µl) versus average steady-state emactuzumab concentration (µg/ml) by dosing cycle for monotherapy (upper plot, P1A, n = 27) and in combination with paclitaxel (lower plot, P1B, n = 23). (B) CSF-1 levels (pg/ml) versus average steady-state emactuzumab concentration (µg/ml) by dosing cycle in monotherapy (upper plot, P1A, n = 27) and in combination with paclitaxel (lower plot, P1B, n = 23). (C) Maximum change from baseline CD163+ skin macrophage levels (high power field count) versus average steady-state emactuzumab concentration (µg/ml) by dosing cycle for monotherapy (left plot, P1A, n = 24) and in combination with paclitaxel (right plot, P1B, n = 21). (D) Maximum change from baseline CSF-1R+ skin macrophage levels (high power field count) versus average steady-state emactuzumab lower concentration (µg/ml) by dosing cycle for monotherapy (left plot, P1A, n = 26) and in combination with paclitaxel (right plot, P1B, n = 20). C1D1, cycle 1 day 1; Cave, average approximate concentration; CD, cluster of differentiation; CSF-1R, colony-stimulating factor receptor; IHC, immunohistochemical; P1A, emactuzumab monotherapy; P1B, emactuzumab + paclitaxel.
Inversely to non-classical monocytes, a rapid and sustained CSF-1 increase in the serum was observed on cycle 1, day 1 after emactuzumab administration, concomitantly with systemic emactuzumab levels. CSF-1 serum levels increased between doses of 100 and 400 mg and tended to plateau at doses ≥400 mg whereas in combination with paclitaxel, CSF-1 levels were more variable (Figure 1B).
Dermal CSF-1R- or CD163-expressing macrophages showed an exposure-related decrease, with a plateau at exposures from an average approximate concentration (Cave) of 100 µg/ml, corresponding to emactuzumab ∼900 mg (Figure 1C and D and supplementary Figure S1, available at Annals of Oncology online). This contrasts with the maximum effects observed for circulating PD markers at lower exposure levels. Review of AEs by dosing cohort disclosed no relationship between dose level and incidence of AEs for monotherapy and combination with paclitaxel (data not shown). Based on the PK/PD relationship in surrogate skin tissue, the OBD was defined at 1000 mg.
Fresh paired tumor biopsies were taken at baseline and on treatment at 4 weeks after two cycles of emactuzumab. We observed a reduction of both CSF-1R+ and CD68+/CD163+ macrophages in tumor tissue at various dose levels for monotherapy and in combination with paclitaxel (Figure 2A–C).
Figure 2.
Emactuzumab monotherapy and in combination with paclitaxel reduced tumor-associated macrophages in the tumor microenvironment. (A) Paired pre- and on-treatment tumor biopsies 4 weeks after treatment from individual patients receiving emactuzumab monotherapy at 200, 400, 600, 900, 1000 and 2000 mg (n = 60) and (B) in combination with a fixed weekly dose of paclitaxel 80 mg/m2 paclitaxel (n = 37). Upper panel: % CSF-1R+ -, lower panel: % area coverage of CD68+CD163+ cells. Median % change from baseline = -67.7% for CSF-1R+ and -60% for CD68+CD163+ TAMs for monotherapy and -45.8% for CSF-1R+ and -48.3% for CD68+/CD163+ TAMs for combination with paclitaxel. (C) Representative IHC images of CD68+CD163+ TAM and CSF-1R+ cells in pre- and on-treatment tumor biopsies 4 weeks after treatment from two patients with primary breast and ovarian cancer, respectively. CD, cluster of differentiation; CSF-1R, colony-stimulating factor receptor; IHC, immunohistochemical; TAMs, tumor-associated macrophages.
The mean (median) percentage change from baseline at the emactuzumab dose of 1000 mg was −48.4% (−67.7%) for CSF-1R+ and −50.2% (−60%) for CD68+/CD163+ TAMs as a single agent and −45.9% (−45.8%) for CSF-1R+ and –42.6% (−48.3%) for CD68+/CD163+ TAMs in combination with paclitaxel, with no statistically significant differences between monotherapy and paclitaxel combination. We did not observe an increase of the CD8+/CD4+ T-cell ratio after depletion TAMs as previously observed (supplementary Figure S3A and B, available at Annals of Oncology online) [5].
To further investigate our proposed mode of action to specifically deplete immunosuppressive M2-like TAMs, we carried out gene expression profiling from paired pre- and on-treatment tumor biopsies for pooled data from both emactuzumab monotherapy and in combination with paclitaxel. We identified 144 down- and 82 up-regulated genes (≥2-fold) in on-treatment biopsies relative to their individual baseline expression in 31 patients (Figure 3A). Subsequently, we mapped the down-regulated genes to gene signatures of different immune cells from peripheral blood mononuclear cells of healthy human individuals. Most of the down-regulated genes after two cycles were overexpressed in peripheral cells of the myeloid lineage, such as monocytes and neutrophils, but not in natural killer cells, CD4+ and CD8+ T cells of the lymphoid lineage. Interestingly, down-regulated genes were also observed in B cells (Figure 3B, left column). However, these down-regulated genes were not B-cell specific but could be mapped to genes associated with general MHC expression and co-stimulation, that is antigen presentation (Figure 3B). Most of the up-regulated genes showed low to no expression in immune cells (Figure 3A).
Figure 3.
Preferential depletion of M2-like tumor-associated macrophages by emactuzumab treatment. (A) Heatmap showing relative expression of genes regulated by emactuzumab ± paclitaxel in paired pre- and on-treatment tumor biopsies 4 weeks after treatment from 31 treated patients. Genes are ordered based on log2 ratio (treated/baseline), from most down-regulated (lowest negative) to most up-regulated (highest positive). (B) Comparison of the emactuzumab ± paclitaxel down-regulated relative gene expression with immune cell gene expression data originating from PBMCs of healthy volunteers using the public Gene Expression Omnibus dataset (GSE60424, https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE60424), 217 of the 222 emactuzumab-regulated genes (left side) were mapped to the GSE60424 dataset (right side). Most of the down-regulated genes were overexpressed in monocytes and/or neutrophils. Emactuzumab down-regulated genes that were shared with lymphoid cells could be mapped to B cells and to their common function in antigen-presenting cells such as HLA genes and CD86. (C) Comparison of the emactuzumab ± paclitaxel down-regulated relative gene expression in paired biopsies (left) with gene expression patterns derived from in vitro-differentiated human macrophages (middle). The color coding of genes from in vitro-differentiated human macrophages represents genes that were expressed above the limit of blank in macrophage population in general and in respective subpopulations such as M2 a/c or M1 [21]. The limit of blank is a value obtained from Affymetrix control probe sets; below this limit, expression is not distinguishable from background noise. Unlike the up-regulated genes, down-regulated genes are typically expressed in macrophages and particularly in the M2-like in vitro-differentiated macrophages. Representative genes among the down-regulated genes that were associated with macrophage or M2-like functions (right). (D) Boxplots showing the scores of TAM signatures derived from single cell sequencing studies of metastatic melanoma, NSCLC and breast cancer in baseline and emactuzumab-treated samples. All three signatures had significantly different scores between baseline and treated samples (raw P < 0.05). The TAM signature from melanoma also showed significance after correcting for multiple testing across genes and signatures (FDR 0.0425). (E) Representative scatter plots showcasing the negative correlation between the emactuzumab PD gene signature and a TAM signature derived from single cell study on metastatic melanoma for TCGA primary bladder and colon cancer. A table listing the Spearman correlation values for all TCGA tumor types is provided in the supplementary Material, available at Annals of Oncology online. Generally, the correlation between these two signatures range from to −0.83 (prostate) to −0.98 (glioma) (supplementary Table S1, available at Annals of Oncology online). CD, cluster of differentiation; FDC, false discovery rate; HLA, human leukocyte antigen; NK, natural killer; NSCLC, non-small-cell lung cancer; PBMC, peripheral blood mononuclear cell; PD, pharmacodynamics; TAMs, tumor-associated macrophages; TCGA, The Cancer Genome Atlas.
We compared the altered gene-to-gene expression profiles generated from in vitro-differentiated human M1- and M2-like macrophage subpopulations [5, 15]. The majority of down-regulated genes were associated with genes that are overexpressed in in vitro-differentiated M2-like macrophages, such as FCGR1A, CSF-1R, MSR1, CD163, C1QA and CD300C [16–18] indicating the specific M2-like TAM depletion by emactuzumab (Figure 3C; for a complete list of genes see supplementary Table S1, available at Annals of Oncology online). Up-regulated genes did not associate with macrophage-related genes (Figure 3C).
We also compared the PD changes in gene expression with published treatment-agnostic TAM signatures derived from single cell sequencing studies of metastatic melanoma [19], non-small-cell lung cancer [20] and breast cancer [21]. We observed that the patient-derived TAM signatures had significantly lower on-treatment scores versus baseline pretreatment samples (P < 0.05) compared with the observed PD gene signature for emactuzumab alone or in combination with paclitaxel (Figure 3D). We also saw a negative correlation of the emactuzumab PD gene signature compared with the melanoma-derived single cell TAM signature [19] and other tumor types from TGCA, such as bladder and colon cancer (Figure 3E). Supplementary Table S1, available at Annals of Oncology online lists the Spearman correlation values for all TCGA tumor types.
Discussion
Emactuzumab, as single agent or in combination with paclitaxel, did not reach an MTD. The safety profile, including asthenia, edema and asymptomatic liver enzyme elevations, was in line with previous reports on CSF-1R-targeting antibodies as monotherapy [7, 22, 23] and the combination of the CSF-1R kinase inhibitor pexidartinib and paclitaxel [24]. Kupffer cells play an important role in the clearance of serum enzymes, including AST, ALT, LDH, and CPK [12]. The observed elevations in serum enzymes are consistent with their delayed clearance secondary to emactuzumab-induced depletion of Kupffer cells. The vast majority of patients remained asymptomatic despite continued treatment at the recommended dose. None of the patients developed liver failure or rhabdomyolysis.
Although clinical activity appeared early in the FDG–PET, with a metabolic ORR of 11%, emactuzumab alone showed no objective RECIST responses, indicating that CSF-1R-blockade has marginal therapeutic benefit in solid tumors. This is consistent with other observations from clinical studies of CSF-1/CSF-1R-targeting therapies in solid tumors [25–27]. Although TAM depletion has been shown to enhance chemotherapy efficacy [28], the ORR for emactuzumab in combination with paclitaxel was lower in the present study (7% across dose cohorts) than seen in clinical studies of paclitaxel alone in higher treatment lines (ranging from 13% to 43% for HER2-negative breast cancer) [29–31] and might be explained by multiple factors (e.g. number of previous treatment lines, dose of emactuzumab, prior taxane therapy).
Our data confirm previous reports on the CSF-1R-dependent reduction of nonclassical peripheral monocytes [5, 27, 32, 33]. We hypothesize that the susceptibility of the described monocyte subsets to emactuzumab is associated with their individual CSF-1R expression levels [5], as these might indicate a respective dependence on CSF-1R-mediated survival signals. Importantly, with the depletion of the non-classical monocytes from the periphery, we did not observe any wound healing defects or other immune system-compromising effects in patients.
Interestingly, the PD plateaus for peripheral blood markers were observed at lower exposures/doses than those for surrogate skin and tumor tissue. Paired skin biopsies were collected as surrogate PD tissue as we identified M2-like macrophages in the skin in previous, unrelated clinical studies (unpublished data). Here, we could confirm an association of the observed PD effect in the skin and in the tumor biopsies for the first time. We conclude that emactuzumab 1000 mg q2w represents the OBD achieving optimal target saturation, resulting in maximal depletion of TAMs in tissue without altering the safety and tolerability profile. Our data highlight that relying exclusively on PD effects in peripheral blood is not sufficient to determine the OBD for agents that do not reach an MTD.
Paired tumor biopsy analysis demonstrated that emactuzumab efficiently reduced CSF-1R+ and CD163+ TAMs, i.e. markers enriched in T-cell-suppressive M2-like macrophages [5]. In addition, gene expression data suggest that emactuzumab predominantly targets immunosuppressive M2-like TAMs as down-regulated genes largely overlapped with in vitro-differentiated immunosuppressive M2-like macrophages [15, 34] and published in vivo TAM signatures [19–21].
Notably, a fraction of CSF-1R+ or CD163+ TAMs survived at and above the OBD for both emactuzumab monotherapy and in combination with paclitaxel, but tumor infiltration by T cells was not altered. We hypothesize that the expression/detection of CSF-1R on macrophages may not determine susceptibility to emactuzumab per se. Similar to circulating monocyte subsets, the CSF-1R expression levels on TAMs may regulate response to CSF-1R blockade [5]. However, IHC has its limitations regarding the sensitive quantification of few cytokine receptor signals. Furthermore, we failed to detect significant regulation of myeloid-derived suppressor cells [35, 36] in the TME after emactuzumab treatment and hence ruled out a compensatory mechanism involving other immunosuppressive myeloid cells (data not shown).
Our study adds to the data that consistently demonstrate the safety of CSF-1/CSF-1R-targeting therapies in dt-GCT and solid tumors. Appropriate and systematic sampling of serial tissue biopsies, even with limited patient numbers, was crucial for demonstrating mode of action and determining the OBD.
We further conclude that emactuzumab specifically depletes immunosuppressive M2-like macrophages but fails to reprogram remaining macrophages into M1-like immunostimulatory TAMs. While the depletion of TAMs with emactuzumab may indeed prime the resulting TME toward a more pro-inflammatory state, the absence of TAMs alone seems insufficient to drive clinically meaningful antitumor immunity. Hence, the combination with cancer immunotherapies and/or immunostimulatory agents may ultimately bolster an improved pro-inflammatory TME and elicit an immune response against the tumor.
Supplementary Material
Acknowledgements
The study was sponsored by F. Hoffmann-La Roche. The authors thank Elizabeth Quackenbush for critical review of the manuscript, Serafino Pantano for helping setting up the macrophage IHC assays and Anna Kiialainen for data analysis. The authors would also like to thank Mark O’Connor and Claire Lavin (Rx Communications, Mold, UK) for medical writing assistance with the preparation of this article, funded by F. Hoffmann-La Roche.
Funding
Funding for this study was provided by F. Hoffmann-La Roche, Basel, Switzerland. No grant numbers apply.
Disclosure
PAC received honoraria from Novartis, Roche, Blueprint Medicines, Amgen; research funding from Novartis, Roche, Lilly, Blueprint Medicine, Bayer, AstraZeneca, Celgene, Plexxikon, Abbvie, Bristol-Myers Squibb, Merck Serono, Merck Sharp & Dohme, Taiho Pharmaceutical, Toray Industries, Transgene; travel grants from Roche, Amgen, Novartis, Bristol-Myers Squibb. AI is an advisory board member and received honoraria and research grant from Roche. CLT acted as an advisory board member for Merck Sharp & Dohme, BMS, Merck Serono, Roche, Amgen, Nanobiotix, GSK. SPD acted as an advisory board member for EMD Serono, Amgen, Nektar, Immune Design, GSK, Merck, Incyte; received travel grants from Adaptimmune, EDM Serono, Nektar; research grants from Nektar, Merck, Incyte, EMD Serono, Deciphera. MAC is a Roche stock owner and sponsor employee. A-MJ is a sponsor employee. CR is a Roche stock owner, sponsor employee and patent holder. RC is a Roche and BMS stock owner and sponsor employee. GM-L is a Roche stock owner and sponsor employee. WJ is a Roche stock owner and sponsor employee. IK is a sponsor employee. C-HO is a Roche stock owner and sponsor employee. CW is a sponsor consultant. KW is a Roche stock owner and sponsor employee. BR is a Roche stock owner and sponsor employee. DR is a sponsor employee. J-PD is a member of advisory boards for Roche, Novartis and MSD; received research grants from MSD, BMS, Genentech and AstraZeneca. J-YB received research support and honoraria from Roche, Daiichi Sankyo, GSK and Novartis; research support from Five Prime. All remaining authors have declared no conflicts of interest.
Note: This study was previously presented in Gomez-Roca et al. [31].
References
- 1. Mantovani A, Sozzani S, Locati M. et al. Macrophage polarization: tumor-associated macrophages as a paradigm for polarized M2 mononuclear phagocytes. Trends Immunol 2002; 23(11): 549–555. [DOI] [PubMed] [Google Scholar]
- 2. Hao N-B, Lü M-H, Fan Y-H. et al. Macrophages in tumor microenvironments and the progression of tumors. Clin Dev Immunol 2012; 2012: 948098. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Stanley ER, Chitu V. CSF-1 receptor signaling in myeloid cells. Cold Spring Harb Perspect Biol 2014; 6(6). doi: 10.1101/cshperspect.a021857. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Yang L, Zhang Y. Tumor-associated macrophages: from basic research to clinical application. J Hematol Oncol 2017; 10(1): 58.. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Ries CH, Cannarile MA, Hoves S. et al. Targeting tumor-associated macrophages with anti-CSF-1R antibody reveals a strategy for cancer therapy. Cancer Cell 2014; 25(6): 846–859. [DOI] [PubMed] [Google Scholar]
- 6. Ries CH, Hoves S, Cannarile MA. et al. CSF-1/CSF-1R targeting agents in clinical development for cancer therapy. Curr Opin Pharmacol 2015; 23: 45–51. [DOI] [PubMed] [Google Scholar]
- 7. Cassier PA, Italiano A, Gomez-Roca CA. et al. CSF1R inhibition with emactuzumab in locally advanced diffuse-type tenosynovial giant cell tumours of the soft tissue: a dose-escalation and dose-expansion phase 1 study. Lancet Oncol 2015; 16(8): 949–956. [DOI] [PubMed] [Google Scholar]
- 8. Common Terminology Criteria for Adverse Events v4.03 (CTCAE). http://evs.nci.nih.gov/ftp1/CTCAE (20 November 2011, date last accessed).
- 9. Komohara Y, Niino D, Ohnishi K. et al. Role of tumor-associated macrophages in hematological malignancies. Pathol Int 2015; 65(4): 170–176. [DOI] [PubMed] [Google Scholar]
- 10. Eisenhauer EA, Therasse P, Bogaerts J. et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009; 45(2): 228–247. [DOI] [PubMed] [Google Scholar]
- 11. Young H, Baum R, Cremerius U. et al. Measurement of clinical and subclinical tumour response using [18F]-fluorodeoxyglucose and positron emission tomography: review and 1999 EORTC recommendations. European Organization for Research and Treatment of Cancer (EORTC) PET Study Group. Eur J Cancer 1999; 35(13): 1773–1782. [DOI] [PubMed] [Google Scholar]
- 12. Radi ZA, Koza-Taylor PH, Bell RR. et al. Increased serum enzyme levels associated with kupffer cell reduction with no signs of hepatic or skeletal muscle injury. Am J Pathol 2011; 179(1): 240–247. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Olingy CE, San Emeterio CL, Ogle ME. et al. Non-classical monocytes are biased progenitors of wound healing macrophages during soft tissue injury. Sci Rep 2017; 7(1): 447. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Mukherjee R, Kanti Barman P, Kumar Thatoi P. et al. Non-classical monocytes display inflammatory features: validation in sepsis and systemic lupus erythematosus. Sci Rep 2015; 5: 13886.. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Pradel LP, Ooi C-H, Romagnoli S. et al. Macrophage susceptibility to emactuzumab (RG7155) treatment. Mol Cancer Ther 2016; 15(12): 3077–3086. [DOI] [PubMed] [Google Scholar]
- 16. Borrego F The CD300 molecules: an emerging family of regulators of the immune system. Blood 2013; 121(11): 1951–1960. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Labonte AC, Sung S-SJ, Jennelle LT. et al. Expression of scavenger receptor-AI promotes alternative activation of murine macrophages to limit hepatic inflammation and fibrosis. Hepatology 2017; 65(1): 32–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Spivia W, Magno PS, Le P. et al. Complement protein C1q promotes macrophage anti-inflammatory M2-like polarization during the clearance of atherogenic lipoproteins. Inflamm Res 2014; 63(10): 885–893. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Tirosh I, Izar B, Prakadan SM. et al. Dissecting the multicellular ecosystem of metastatic melanoma by single-cell RNA-seq. Science 2016; 352(6282): 189–196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Lavin Y, Kobayashi S, Leader A. et al. Innate immune landscape in early lung adenocarcinoma by paired single-cell analyses. Cell. 2017; 169(4): 750–765.e17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Chung W, Eum HH, Lee H-O. et al. Single-cell RNA-seq enables comprehensive tumour and immune cell profiling in primary breast cancer. Nat Comms 2017; 8: 15081.. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Bendell JC, Tolcher AW, Jones S. et al. A phase 1 study of ARRY-382, an oral inhibitor of colony-stimulating factor-1 receptor (CSF1R), in patients with advanced or metastatic cancers. Mol Cancer Ther 2013; 12(11). [Google Scholar]
- 23. Hambleton J, Zhou L, Rogers S. et al. A phase 1 study of FPA008, an anti-colony stimulating factor 1 receptor (anti-CSF1R) antibody in healthy volunteers and subjects with rheumatoid arthritis (RA): preliminary results. Arthritis Rheumatol 2014; 66: S657–S657. [Google Scholar]
- 24. Rugo HS, Sharma N, Reebel L, et al. Phase Ib study of Plx3397, a Csf1r inhibitor, and paclitaxel in patients with advanced solid tumors. Ann Oncol 2014; 25: iv148. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Cannarile MA, Weisser M, Jacob W. et al. Colony-stimulating factor 1 receptor (CSF1R) inhibitors in cancer therapy. J Immunother Cancer 2017; 5(1): 53.. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Papadopoulos KP, Gluck L, Martin LP. et al. First-in-human study of AMG 820, a monoclonal anti-colony-stimulating factor 1 receptor antibody, in patients with advanced solid tumors. Clin Cancer Res 2017; 23(19): 5703–5710. [DOI] [PubMed] [Google Scholar]
- 27. Wainberg ZA, Piha-Paul SA, Luke J. et al. First-in-human phase 1 dose escalation and expansion of a novel combination, anti–CSF-1 receptor (cabiralizumab) plus anti–PD-1 (nivolumab), in patients with advanced solid tumors. Presented at Society for Immunotherapy of Cancer 2018; 10.13140/RG. 2.2.28962.53443. [Google Scholar]
- 28. Salvagno C, Ciampricotti M, Tuit S. et al. Therapeutic targeting of macrophages enhances chemotherapy efficacy by unleashing type I interferon response. Nat Cell Biol 2019; 21(4): 511–521. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Brufsky AM, Hurvitz S, Perez E. et al. RIBBON-2: a randomized, double-blind, placebo-controlled, phase III trial evaluating the efficacy and safety of bevacizumab in combination with chemotherapy for second-line treatment of human epidermal growth factor receptor 2-negative metastatic breast cancer. J Clin Oncol 2011; 29(32): 4286–4293. [DOI] [PubMed] [Google Scholar]
- 30. Decker T, Overkamp F, Rösel S. et al. A randomized phase II study of paclitaxel alone versus paclitaxel plus sorafenib in second- and third-line treatment of patients with HER2-negative metastatic breast cancer (PASO). BMC Cancer 2017; 17(1): 499.. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Gradishar WJ, Tjulandin S, Davidson N. et al. Phase III trial of nanoparticle albumin-bound paclitaxel compared with polyethylated castor oil-based paclitaxel in women with breast cancer. J Clin Oncol 2005; 23(31): 7794–7803. [DOI] [PubMed] [Google Scholar]
- 32. Gomez-Roca CA, Cassier PA, Italiano A. et al. Phase I study of RG7155, a novel anti-CSF1R antibody, in patients with advanced/metastatic solid tumors. J Clin Oncol 2015; 33(Suppl 15): 3005. [Google Scholar]
- 33. Anthony SP, Puzanov I, Lin PS. et al. Pharmacodynamic activity demonstrated in phase I for PLX3397, a selective inhibitor of FMS and Kit. J Clin Oncol 2011; 29(Suppl 15): 3093. [Google Scholar]
- 34. Pradel LP, Franke A, Ries CH. Effects of IL-10 and Th 2 cytokines on human Mphi phenotype and response to CSF1R inhibitor. J Leukoc Biol 2018; 103(3): 545–558. [DOI] [PubMed] [Google Scholar]
- 35. Youn J-I, Collazo M, Shalova IN. et al. Characterization of the nature of granulocytic myeloid-derived suppressor cells in tumor-bearing mice. J Leukoc Biol 2012; 91(1): 167–181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Abbas AR, Baldwin D, Ma Y. et al. Immune response in silico (IRIS): immune-specific genes identified from a compendium of microarray expression data. Genes Immun 2005; 6(4): 319–331. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.